includes/api/ApiParse.php

   1 <?php
   2 /**
   3  *
   4  *
   5  * Created on Dec 01, 2007
   6  *
   7  * Copyright © 2007 Yuri Astrakhan <Firstname><Lastname>@gmail.com
   8  *
   9  * This program is free software; you can redistribute it and/or modify
  10  * it under the terms of the GNU General Public License as published by
  11  * the Free Software Foundation; either version 2 of the License, or
  12  * (at your option) any later version.
  13  *
  14  * This program is distributed in the hope that it will be useful,
  15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  17  * GNU General Public License for more details.
  18  *
  19  * You should have received a copy of the GNU General Public License along
  20  * with this program; if not, write to the Free Software Foundation, Inc.,
  21  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  22  * http://www.gnu.org/copyleft/gpl.html
  23  *
  24  * @file
  25  */
  26
  27 if ( !defined( 'MEDIAWIKI' ) ) {
  28         // Eclipse helper - will be ignored in production
  29         require_once( "ApiBase.php" );
  30 }
  31
  32 /**
  33  * @ingroup API
  34  */
  35 class ApiParse extends ApiBase {
  36
  37         private $section;
  38
  39         public function __construct( $main, $action ) {
  40                 parent::__construct( $main, $action );
  41         }
  42
  43         public function execute() {
  44                 // The data is hot but user-dependent, like page views, so we set vary cookies
  45                 $this->getMain()->setCacheMode( 'anon-public-user-private' );
  46
  47                 // Get parameters
  48                 $params = $this->extractRequestParams();
  49                 $text = $params['text'];
  50                 $title = $params['title'];
  51                 $page = $params['page'];
  52                 $pageid = $params['pageid'];
  53                 $oldid = $params['oldid'];
  54
  55                 if ( !is_null( $page ) && ( !is_null( $text ) || $title != 'API' ) ) {
  56                         $this->dieUsage( 'The page parameter cannot be used together with the text and title parameters', 'params' );
  57                 }
  58                 $prop = array_flip( $params['prop'] );
  59
  60                 if ( isset( $params['section'] ) ) {
  61                         $this->section = $params['section'];
  62                 } else {
  63                         $this->section = false;
  64                 }
  65
  66                 // The parser needs $wgTitle to be set, apparently the
  67                 // $title parameter in Parser::parse isn't enough *sigh*
  68                 global $wgParser, $wgUser, $wgTitle, $wgEnableParserCache, $wgLang;
  69
  70                 // Currently unnecessary, code to act as a safeguard against any change in current behaviour of uselang breaks
  71                 $oldLang = null;
  72                 if ( isset( $params['uselang'] ) && $params['uselang'] != $wgLang->getCode() ) {
  73                         $oldLang = $wgLang; // Backup wgLang
  74                         $wgLang = Language::factory( $params['uselang'] );
  75                 }
  76
  77                 $popts = new ParserOptions();
  78                 $popts->setTidy( true );
  79                 $popts->enableLimitReport( !$params['disablepp'] );
  80                 $redirValues = null;
  81                 if ( !is_null( $oldid ) || !is_null( $pageid ) || !is_null( $page ) ) {
  82                         if ( !is_null( $oldid ) ) {
  83                                 // Don't use the parser cache
  84                                 $rev = Revision::newFromID( $oldid );
  85                                 if ( !$rev ) {
  86                                         $this->dieUsage( "There is no revision ID $oldid", 'missingrev' );
  87                                 }
  88                                 if ( !$rev->userCan( Revision::DELETED_TEXT ) ) {
  89                                         $this->dieUsage( "You don't have permission to view deleted revisions", 'permissiondenied' );
  90                                 }
  91
  92                                 $titleObj = $rev->getTitle();
  93
  94                                 $wgTitle = $titleObj;
  95
  96                                 // If for some reason the "oldid" is actually the current revision, it may be cached
  97                                 if ( $titleObj->getLatestRevID() == $oldid ) {
  98                                         $p_result = false;
  99                                         $pcache = ParserCache::singleton();
 100
 101                                         $articleObj = new Article( $titleObj );
 102
 103                                         if ( $wgEnableParserCache ) {
 104                                                 $p_result = $pcache->get( $articleObj, $popts );
 105                                         }
 106                                         if ( !$p_result ) {
 107                                                 $text = $rev->getText( Revision::FOR_THIS_USER );
 108                                                 $p_result = $wgParser->parse( $text, $titleObj, $popts );
 109
 110                                                 if ( $wgEnableParserCache ) {
 111                                                         $pcache->save( $p_result, $articleObj, $popts );
 112                                                 }
 113                                         }
 114                                 } else {
 115                                         $text = $rev->getText( Revision::FOR_THIS_USER );
 116
 117                                         $wgTitle = $titleObj;
 118
 119                                         if ( $this->section !== false ) {
 120                                                 $text = $this->getSectionText( $text, 'r' . $rev );
 121                                         }
 122
 123                                         $p_result = $wgParser->parse( $text, $titleObj, $popts );
 124                                 }
 125                         } else {
 126                                 if ( !is_null ( $pageid ) ) {
 127                                         $titleObj = Title::newFromID( $pageid );
 128
 129                                         if ( !$titleObj ) {
 130                                                 $this->dieUsageMsg( array( 'nosuchpageid', $pageid ) );
 131                                         }
 132                                 } else {
 133                                         if ( $params['redirects'] ) {
 134                                                 $req = new FauxRequest( array(
 135                                                         'action' => 'query',
 136                                                         'redirects' => '',
 137                                                         'titles' => $page
 138                                                 ) );
 139                                                 $main = new ApiMain( $req );
 140                                                 $main->execute();
 141                                                 $data = $main->getResultData();
 142                                                 $redirValues = @$data['query']['redirects'];
 143                                                 $to = $page;
 144                                                 foreach ( (array)$redirValues as $r ) {
 145                                                         $to = $r['to'];
 146                                                 }
 147                                         } else {
 148                                                 $to = $page;
 149                                         }
 150                                         $titleObj = Title::newFromText( $to );
 151                                         if ( !$titleObj ) {
 152                                                 $this->dieUsage( "The page you specified doesn't exist", 'missingtitle' );
 153                                         }
 154                                 }
 155                                 $wgTitle = $titleObj;
 156
 157                                 $articleObj = new Article( $titleObj );
 158                                 if ( isset( $prop['revid'] ) ) {
 159                                         $oldid = $articleObj->getRevIdFetched();
 160                                 }
 161
 162                                 if ( $this->section !== false ) {
 163                                         $text = $this->getSectionText( $text, !is_null ( $pageid )
 164                                                         ? 'page id ' . $pageid : $titleObj->getText() );
 165
 166                                         $p_result = $wgParser->parse( $text, $titleObj, $popts );
 167                                 } else {
 168                                         // Try the parser cache first
 169                                         $p_result = false;
 170                                         $pcache = ParserCache::singleton();
 171                                         if ( $wgEnableParserCache ) {
 172                                                 $p_result = $pcache->get( $articleObj, $popts );
 173                                         }
 174                                         if ( !$p_result ) {
 175                                                 $p_result = $wgParser->parse( $articleObj->getContent(), $titleObj, $popts );
 176
 177                                                 if ( $wgEnableParserCache ) {
 178                                                         $pcache->save( $p_result, $articleObj, $popts );
 179                                                 }
 180                                         }
 181                                 }
 182                         }
 183                 } else {
 184                         $titleObj = Title::newFromText( $title );
 185                         if ( !$titleObj ) {
 186                                 $titleObj = Title::newFromText( 'API' );
 187                         }
 188                         $wgTitle = $titleObj;
 189
 190                         if ( $this->section !== false ) {
 191                                 $text = $this->getSectionText( $text, $titleObj->getText() );
 192                         }
 193
 194                         if ( $params['pst'] || $params['onlypst'] ) {
 195                                 $text = $wgParser->preSaveTransform( $text, $titleObj, $wgUser, $popts );
 196                         }
 197                         if ( $params['onlypst'] ) {
 198                                 // Build a result and bail out
 199                                 $result_array['text'] = array();
 200                                 $this->getResult()->setContent( $result_array['text'], $text );
 201                                 $this->getResult()->addValue( null, $this->getModuleName(), $result_array );
 202                                 return;
 203                         }
 204                         $p_result = $wgParser->parse( $text, $titleObj, $popts );
 205                 }
 206
 207                 // Return result
 208                 $result = $this->getResult();
 209                 $result_array = array();
 210                 if ( $params['redirects'] && !is_null( $redirValues ) ) {
 211                         $result_array['redirects'] = $redirValues;
 212                 }
 213
 214                 if ( isset( $prop['text'] ) ) {
 215                         $result_array['text'] = array();
 216                         $result->setContent( $result_array['text'], $p_result->getText() );
 217                 }
 218
 219                 if ( !is_null( $params['summary'] ) ) {
 220                         $result_array['parsedsummary'] = array();
 221                         $result->setContent( $result_array['parsedsummary'], $wgUser->getSkin()->formatComment( $params['summary'], $titleObj ) );
 222                 }
 223
 224                 if ( isset( $prop['langlinks'] ) ) {
 225                         $result_array['langlinks'] = $this->formatLangLinks( $p_result->getLanguageLinks() );
 226                 }
 227                 if ( isset( $prop['languageshtml'] ) ) {
 228                         $languagesHtml = $this->languagesHtml( $p_result->getLanguageLinks() );
 229                         $result_array['languageshtml'] = array();
 230                         $result->setContent( $result_array['languageshtml'], $languagesHtml );
 231                 }
 232                 if ( isset( $prop['categories'] ) ) {
 233                         $result_array['categories'] = $this->formatCategoryLinks( $p_result->getCategories() );
 234                 }
 235                 if ( isset( $prop['categorieshtml'] ) ) {
 236                         $categoriesHtml = $this->categoriesHtml( $p_result->getCategories() );
 237                         $result_array['categorieshtml'] = array();
 238                         $result->setContent( $result_array['categorieshtml'], $categoriesHtml );
 239                 }
 240                 if ( isset( $prop['links'] ) ) {
 241                         $result_array['links'] = $this->formatLinks( $p_result->getLinks() );
 242                 }
 243                 if ( isset( $prop['templates'] ) ) {
 244                         $result_array['templates'] = $this->formatLinks( $p_result->getTemplates() );
 245                 }
 246                 if ( isset( $prop['images'] ) ) {
 247                         $result_array['images'] = array_keys( $p_result->getImages() );
 248                 }
 249                 if ( isset( $prop['externallinks'] ) ) {
 250                         $result_array['externallinks'] = array_keys( $p_result->getExternalLinks() );
 251                 }
 252                 if ( isset( $prop['sections'] ) ) {
 253                         $result_array['sections'] = $p_result->getSections();
 254                 }
 255
 256                 if ( isset( $prop['displaytitle'] ) ) {
 257                         $result_array['displaytitle'] = $p_result->getDisplayTitle() ?
 258                                                         $p_result->getDisplayTitle() :
 259                                                         $titleObj->getPrefixedText();
 260                 }
 261
 262                 if ( isset( $prop['headitems'] ) || isset( $prop['headhtml'] ) ) {
 263                         $out = new OutputPage;
 264                         $out->addParserOutputNoText( $p_result );
 265                         $userSkin = $wgUser->getSkin();
 266                 }
 267
 268                 if ( isset( $prop['headitems'] ) ) {
 269                         $headItems = $this->formatHeadItems( $p_result->getHeadItems() );
 270
 271                         $userSkin->setupUserCss( $out );
 272                         $css = $this->formatCss( $out->buildCssLinksArray() );
 273
 274                         $scripts = array( $out->getHeadScripts( $userSkin ) );
 275
 276                         $result_array['headitems'] = array_merge( $headItems, $css, $scripts );
 277                 }
 278
 279                 if ( isset( $prop['headhtml'] ) ) {
 280                         $result_array['headhtml'] = array();
 281                         $result->setContent( $result_array['headhtml'], $out->headElement( $userSkin ) );
 282                 }
 283
 284                 if ( isset( $prop['iwlinks'] ) ) {
 285                         $result_array['iwlinks'] = $this->formatIWLinks( $p_result->getInterwikiLinks() );
 286                 }
 287
 288                 if ( !is_null( $oldid ) ) {
 289                         $result_array['revid'] = intval( $oldid );
 290                 }
 291
 292                 $result_mapping = array(
 293                         'redirects' => 'r',
 294                         'langlinks' => 'll',
 295                         'categories' => 'cl',
 296                         'links' => 'pl',
 297                         'templates' => 'tl',
 298                         'images' => 'img',
 299                         'externallinks' => 'el',
 300                         'iwlinks' => 'iw',
 301                         'sections' => 's',
 302                         'headitems' => 'hi',
 303                 );
 304                 $this->setIndexedTagNames( $result_array, $result_mapping );
 305                 $result->addValue( null, $this->getModuleName(), $result_array );
 306
 307                 if ( !is_null( $oldLang ) ) {
 308                         $wgLang = $oldLang; // Reset $wgLang to $oldLang
 309                 }
 310         }
 311
 312         private function getSectionText( $text, $what ) {
 313                 global $wgParser;
 314                 $text = $wgParser->getSection( $text, $this->section, false );
 315                 if ( $text === false ) {
 316                         $this->dieUsage( "There is no section {$this->section} in " . $what, 'nosuchsection' );
 317                 }
 318                 return $text;
 319         }
 320
 321         private function formatLangLinks( $links ) {
 322                 $result = array();
 323                 foreach ( $links as $link ) {
 324                         $entry = array();
 325                         $bits = explode( ':', $link, 2 );
 326                         $entry['lang'] = $bits[0];
 327                         $this->getResult()->setContent( $entry, $bits[1] );
 328                         $result[] = $entry;
 329                 }
 330                 return $result;
 331         }
 332
 333         private function formatCategoryLinks( $links ) {
 334                 $result = array();
 335                 foreach ( $links as $link => $sortkey ) {
 336                         $entry = array();
 337                         $entry['sortkey'] = $sortkey;
 338                         $this->getResult()->setContent( $entry, $link );
 339                         $result[] = $entry;
 340                 }
 341                 return $result;
 342         }
 343
 344         private function categoriesHtml( $categories ) {
 345                 global $wgOut, $wgUser;
 346                 $wgOut->addCategoryLinks( $categories );
 347                 $sk = $wgUser->getSkin();
 348                 return $sk->getCategories();
 349         }
 350
 351         private function languagesHtml( $languages ) {
 352                 global $wgOut, $wgUser;
 353                 $wgOut->setLanguageLinks( $languages );
 354                 $sk = $wgUser->getSkin();
 355                 return $sk->otherLanguages();
 356         }
 357
 358         private function formatLinks( $links ) {
 359                 $result = array();
 360                 foreach ( $links as $ns => $nslinks ) {
 361                         foreach ( $nslinks as $title => $id ) {
 362                                 $entry = array();
 363                                 $entry['ns'] = $ns;
 364                                 $this->getResult()->setContent( $entry, Title::makeTitle( $ns, $title )->getFullText() );
 365                                 if ( $id != 0 ) {
 366                                         $entry['exists'] = '';
 367                                 }
 368                                 $result[] = $entry;
 369                         }
 370                 }
 371                 return $result;
 372         }
 373
 374         private function formatIWLinks( $iw ) {
 375                 $result = array();
 376                 foreach ( $iw as $prefix => $titles ) {
 377                         foreach ( array_keys( $titles ) as $title ) {
 378                                 $entry = array();
 379                                 $entry['prefix'] = $prefix;
 380
 381                                 $title = Title::newFromText( "{$prefix}:{$title}" );
 382                                 if ( $title ) {
 383                                         $entry['url'] = $title->getFullURL();
 384                                 }
 385
 386                                 $this->getResult()->setContent( $entry, $title->getFullText() );
 387                                 $result[] = $entry;
 388                         }
 389                 }
 390                 return $result;
 391         }
 392
 393         private function formatHeadItems( $headItems ) {
 394                 $result = array();
 395                 foreach ( $headItems as $tag => $content ) {
 396                         $entry = array();
 397                         $entry['tag'] = $tag;
 398                         $this->getResult()->setContent( $entry, $content );
 399                         $result[] = $entry;
 400                 }
 401                 return $result;
 402         }
 403
 404         private function formatCss( $css ) {
 405                 $result = array();
 406                 foreach ( $css as $file => $link ) {
 407                         $entry = array();
 408                         $entry['file'] = $file;
 409                         $this->getResult()->setContent( $entry, $link );
 410                         $result[] = $entry;
 411                 }
 412                 return $result;
 413         }
 414
 415         private function setIndexedTagNames( &$array, $mapping ) {
 416                 foreach ( $mapping as $key => $name ) {
 417                         if ( isset( $array[$key] ) ) {
 418                                 $this->getResult()->setIndexedTagName( $array[$key], $name );
 419                         }
 420                 }
 421         }
 422
 423         public function getAllowedParams() {
 424                 return array(
 425                         'title' => array(
 426                                 ApiBase::PARAM_DFLT => 'API',
 427                         ),
 428                         'text' => null,
 429                         'summary' => null,
 430                         'page' => null,
 431                         'pageid' => null,
 432                         'redirects' => false,
 433                         'oldid' => null,
 434                         'prop' => array(
 435                                 ApiBase::PARAM_DFLT => 'text|langlinks|categories|links|templates|images|externallinks|sections|revid|displaytitle',
 436                                 ApiBase::PARAM_ISMULTI => true,
 437                                 ApiBase::PARAM_TYPE => array(
 438                                         'text',
 439                                         'langlinks',
 440                                         'languageshtml',
 441                                         'categories',
 442                                         'categorieshtml',
 443                                         'links',
 444                                         'templates',
 445                                         'images',
 446                                         'externallinks',
 447                                         'sections',
 448                                         'revid',
 449                                         'displaytitle',
 450                                         'headitems',
 451                                         'headhtml',
 452                                         'iwlinks',
 453                                 )
 454                         ),
 455                         'pst' => false,
 456                         'onlypst' => false,
 457                         'uselang' => null,
 458                         'section' => null,
 459                         'disablepp' => false,
 460                 );
 461         }
 462
 463         public function getParamDescription() {
 464                 $p = $this->getModulePrefix();
 465                 return array(
 466                         'text' => 'Wikitext to parse',
 467                         'summary' => 'Summary to parse',
 468                         'redirects' => "If the {$p}page parameter is set to a redirect, resolve it",
 469                         'title' => 'Title of page the text belongs to',
 470                         'page' => "Parse the content of this page. Cannot be used together with {$p}text and {$p}title",
 471                         'pageid' => "Parse the content of this page. Overrides {$p}page",
 472                         'oldid' => "Parse the content of this revision. Overrides {$p}page and {$p}pageid",
 473                         'prop' => array(
 474                                 'Which pieces of information to get',
 475                                 ' text           - Gives the parsed text of the wikitext',
 476                                 ' langlinks      - Gives the language links in the parsed wikitext',
 477                                 ' categories     - Gives the categories in the parsed wikitext',
 478                                 ' categorieshtml - Gives the HTML version of the categories',
 479                                 ' languageshtml  - Gives the HTML version of the language links',
 480                                 ' links          - Gives the internal links in the parsed wikitext',
 481                                 ' templates      - Gives the templates in the parsed wikitext',
 482                                 ' images         - Gives the images in the parsed wikitext',
 483                                 ' externallinks  - Gives the external links in the parsed wikitext',
 484                                 ' sections       - Gives the sections in the parsed wikitext',
 485                                 ' revid          - Adds the revision ID of the parsed page',
 486                                 ' displaytitle   - Adds the title of the parsed wikitext',
 487                                 ' headitems      - Gives items to put in the <head> of the page',
 488                                 ' headhtml       - Gives parsed <head> of the page',
 489                                 ' iwlinks        - Gives interwiki links in the parsed wikitext',
 490                                 'NOTE: Section tree is only generated if there are more than 4 sections, or if the __TOC__ keyword is present'
 491                         ),
 492                         'pst' => array(
 493                                 'Do a pre-save transform on the input before parsing it',
 494                                 'Ignored if page, pageid or oldid is used'
 495                         ),
 496                         'onlypst' => array(
 497                                 'Do a pre-save transform (PST) on the input, but don\'t parse it',
 498                                 'Returns the same wikitext, after a PST has been applied. Ignored if page, pageid or oldid is used'
 499                         ),
 500                         'uselang' => 'Which language to parse the request in',
 501                         'section' => 'Only retrieve the content of this section number',
 502                         'disablepp' => 'Disable the PP Report from the parser output',
 503                 );
 504         }
 505
 506         public function getDescription() {
 507                 return 'This module parses wikitext and returns parser output';
 508         }
 509
 510         public function getPossibleErrors() {
 511                 return array_merge( parent::getPossibleErrors(), array(
 512                         array( 'code' => 'params', 'info' => 'The page parameter cannot be used together with the text and title parameters' ),
 513                         array( 'code' => 'missingrev', 'info' => 'There is no revision ID oldid' ),
 514                         array( 'code' => 'permissiondenied', 'info' => 'You don\'t have permission to view deleted revisions' ),
 515                         array( 'code' => 'missingtitle', 'info' => 'The page you specified doesn\'t exist' ),
 516                         array( 'code' => 'nosuchsection', 'info' => 'There is no section sectionnumber in page' ),
 517                         array( 'nosuchpageid' ),
 518                 ) );
 519         }
 520
 521         protected function getExamples() {
 522                 return array(
 523                         'api.php?action=parse&text={{Project:Sandbox}}'
 524                 );
 525         }
 526
 527         public function getVersion() {
 528                 return __CLASS__ . ': $Id$';
 529         }
 530 }